R&D: why failure is necessary

The US government’s reclassification of R&D as a sexy investment, not an iffy expense, is foolish.

In July this year, US government statisticians redefined research and development (R&D) as akin to capital goods. R&D spending, they said, amounts to fixed assets, or investment; it should now count toward US gross domestic product (GDP) (1). And last week, this change showed up in the US Federal Reserve’s data on flows of finance in America, helpfully recording inflated asset values and lower debts.

This all seems innocuous enough. After all, R&D is just a down-payment on the future, so why not redefine it as a capital item rather than an expense?

But things are not that simple. R&D is essential, not just to economic growth, but also to the self-respect of society – for without scientific curiosity and technological adventure, the world would be a static, drab place. However, every successful outcome of R&D is usually accompanied by fruitless scientific enquiry, malfunctioning prototypes, and dismal market failures: take forlorn phone maker Blackberry today, or GlaxoSmithKline’s diabetes drug, Avandia, which was dropped in 2011 because it caused heart problems. In general, therefore, R&D pays back rather rarely (2).

Of course, with firms making established goods or services, the caprice of the market ensures that bankruptcy or urgent state bailouts are always possible – think British Leyland in the 1970s, or General Motors after 2008. Yet the direct, immediate production of profit tends to prevail. But with R&D, profit prospects are intrinsically low, and usually lie years away.

When it isn’t trying something entirely new, the brave innovating firm confronts other suppliers – their experience, sunk costs, distribution channels and brand reputation. Also, as Alexander Fleming’s chance discovery of penicillin famously confirms, the innovating firm often needs some real luck. Altogether, then, R&D tends only to succeed after repeated efforts. Turn a profit from making another packet of crisps? Easy enough. Convince airlines of your brand new jet’s airworthiness? As Boeing’s faulty 787 Dreamliner has shown, that’s tough.

No wonder number-crunchers at the Bureau of Economic Analysis (BEA), Department of Commerce, Washington, ask themselves: ‘What about research that leads to nothing?’ Well, the BEA believes that R&D is ‘symmetric with unsuccessful investments in physical capital including unsuccessful oil wells’, as well as with equipment that’s faulty, prematurely scrapped because of new technology, or sidelined because of changes in demand or business closures.

Yet R&D is not ‘symmetric’ with redundant equipment. When activated by labour, equipment usually aids the accumulation of capital in direct fashion. Realising, through sales, the value and surplus value in familiar use values may prove hard; but redundant equipment, products and services are the exception, not the rule.

With R&D, by contrast, failed experiments, though unavoidable and frequently very revealing, tend to be the rule.

So what made the BEA rethink R&D? Discount its excuse that rejigging squares US national accounts with international practice. These days, America cares little what others think and do. Forget, too, the tiny boost that the revision gives to growth statistics (3). More interesting is this: the BEA has actually computed R&D as an investment since 1994 – quietly, as a ‘satellite’ from its main accounts.

So why did it move now?

Federal Reserve Board chairman Ben Bernanke revealed why in 2011. ‘We will be more likely to promote innovative activity’, he remarked, ‘if we are able to measure it more effectively’. Yet US R&D has long been measured – and since the Crash of 2008, spending on it by business has actually declined somewhat, while spending by the federal government and universities has failed to compensate for that:

US business, federal and higher education spending on R&D, 2000-2011 (billions of US dollars, constant 2005 value)

If, to gain some sense of America’s overall commitment of resources to R&D, we look at the percentage of GDP taken by it (research ‘intensity’), there’s a similar tale. With gross (private and public) R&D spending, and that on ‘basic’, scientific research, commitments seem to have faded, slightly, since 2009:

US gross expenditures on R&D and basic research, as a percentage of GDP, 2000-2011

The outcomes of R&D are necessarily uncertain. So even in the years when absolute R&D spending, research intensity and the intensity of basic research all rose, that didn’t necessarily mean more or better innovations. Indeed, a number of commentators in the US have rightly concluded that innovation there is decelerating, not accelerating (4).

Despite its innovation slowdown, American capitalism is not about to up its outlays on R&D. Nor is it about to improve the quality, direction and management of R&D. Nor, finally, are US researchers about to regain the ambitious sense of purpose that marked the Apollo space programme.

What happens instead is that bean-counters in Washington, following Bernanke, purr that R&D succeeds as often as drilling for oil does – a topical analogy, given the rise of shale oil and gas in the US, but not a convincing one. In the same way, the BEA has also announced that it will, henceforth, lump R&D together with films, fiction and fine art: in its best of all possible worlds, everything counts as an investment.

It is easier to confer status on R&D by a keystroke in President Obama’s spreadsheet than it is by a nationwide political and economic drive to reinvigorate innovation. America’s new ‘recognition’ of R&D is entirely complacent. It has nothing in common with the questing, pioneering and tenacious spirit that has always been, and remains, characteristic of genuine R&D.


  • (will not be published)